Thursday 19 September 2013

How to mine the consensus view

The latest consensus forecasts compiled by the NZ Institute of Economic Research (NZIER), were published this week, and it reminded me that a few years back I prepared a module on economics for directors, as part of the Institute of Directors' training courses. The module included some discussion on how to get the best of out these forecasts, from a practical business perspective, which I thought might be worth resurrecting.

First of all, for those unfamiliar with them, these consensus forecasts are the average forecast for a bunch of things, calculated from the forecasts of 10 different forecasters (two public sector, Treasury and the Reserve Bank, and the rest from a range of private sector financial institutions, including all the main banks, plus the NZIER itself).

The averaging is the first benefit of this exercise - there's evidence that the best forecasts, over time, are these averaged consensus ones. Individual forecasters can be all over the place - for example, as I noted earlier, there are currently quite different views among the big banks about how the Australian economy will fare - but the consensus tends to give a more reliable signal. If you're looking for some numbers to plug into your strategic planning, these are as good as any.

Next, the most important element is what the consensus is saying about the outlook for the economy as a whole. Have a look at Table 1. You can see in the columns headed 'Sept-2013 survey', that GDP growth is expected to be 2.6% in the year to March '14, 3.0% in the year to March '15, and 2.3% in the year to March '16. From a business point of view, is this good, bad or indifferent? It's reasonably good. It's enough, for example, as you can see lower in the table when you look at the forecasts for employment growth and for the unemployment rate, to lead to businesses hiring more people each year, on a scale enough to lead to a gradual modest decline in the number of people unemployed.

The next thing I tend to look at it is the change (if any) from the previous consensus. Table 1 helpfully compares the latest set (collated September) with the previous set (collected in June). Overall, this time round, there's no significant change. Sometimes, though, you'll find that economic prospects have improved or deteriorated quite a lot over the space of a quarter. When it happens, and there's been a positive or negative surprise, it's a useful thing to tuck away for planning or risk management purposes.

Next question to ask yourself is, what is this forecast economic growth principally based on? What's the biggest moving part? This time round, it's the very large forecast rise in 'Fixed investment - residential', or housebuilding in other words, which in turn reflects the scale of the Canterbury rebuild. That sounds like a reasonably high-probability bankable proposition: we know the rebuild is going to have to happen.

Other times, though, the forecasts may be based on expected strong growth in export markets, or on growth in government spending, and you need to know that's the basis of current expectations, especially if you have a different view (for example you're finding it tough in export markets at the current level of the Kiwi dollar but forecasters seem to be picking buoyant export markets) or there's a sudden shift in the winds (for example, a change in economic policy).

The consensus forecasts also show you the range of the forecasts: for each variable you can see the most optimistic and the most pessimistic view amongst the ten forecasters. The graphs on pp2-3 are the easiest way of seeing the range of views, and the numbers themselves are in Table 3. Again, it's very helpful to see the spread of views, as it gives you some feel about the degree of uncertainty ahead.

Looking at the GDP graph (below), you can see, for example, that there isn't a single forecaster predicting a recession on the horizon over the next three years. That's not to say they're going to be right - a squall of some kind could materialise out of the blue - but it's a reassuring feature of the forecasts nonetheless. On p3 you can see a similar consensus about inflation - everyone believes the low point is behind us, everyone believes inflation will pick up, but everyone also believes it won't breach the Reserve Bank's 3% maximum.


There's a similar consensus about short term interest rates: everyone expects they're on the way up (implicitly, they're expecting the Reserve Bank to be reacting to that anticipated rise in inflation). Ditto long term interest rates: if you're looking at that 'fixed or floating' interest rate decision, then Table 3 gives you some numbers for expected short and long term interest rates to go into your calculations.

There's also consensus about the overall value of the Kiwi dollar, but personally I don't pay too much to this forecast. For one thing, forecasting exchange rates tends to be an especially unreliable process in the first place (even on a consensus basis). And for another, exchange rate forecasters tend to make the same forecast over and over again: when a currency has been rising, they expect it to rise a bit more, but then decline (the picture, yet again, in these forecasts), and when a currency has been falling, they expect it to fall a bit more, and then rise.

The best that can be said about that shape of forecast is that forecasters have some idea of a long-term 'right' value for the Kiwi dollar, and the further the actual exchange rate has moved away from it, the more they expect it to turn back towards it next time. But you could also less charitably describe the forecasting as purely mechanical.

There's quite a lot of mutual agreement among the forecasters: is there anything where there's a wide spread of views? This time, there's some disagreement about the outlook for exports (graph, p2), but I don't think it means much for businesses. The range of views on exports is (I reckon) down to different views of the one-off impact of the past drought, and doesn't mean anything much for the future export outlook.

You'll find, though, that sometimes the uncertainties are more meaningful. When I first did this exercise for the Institute of Directors, there was considerable uncertainty about the outlook for wages (the economy was growing quite strongly at the time): one possible business response would be to take some of that potential cost volatility out of your business by agreeing on earlier than usual pay increases, or perhaps for longer terms.

There are other ways for folks in business to make productive use of these forecasts. One final one: let's take the year to March '15 as an example. In that year, the consensus forecast is for real GDP growth to be 3%, and the consensus forecast for inflation is for 2.2%. Implicitly, the forecast for growth in nominal GDP (i.e. GDP in everyday dollars) is 5.3% (there's a slight compounding effect that makes it 5.3% rather than the 5.2% you get by adding 3% and 2.2%). So what sort of number have you got in your sales forecasts for that year?

If it's less than 5.3%, you're saying that you're planning to do less well than the average other guy. If it's more than 5.3%, you're expecting to do rather better than firms as a whole. There might be good reasons for both views - but it's also good to know what judgement call you're implicitly making.

No comments:

Post a Comment

Hi - sorry about the Captcha step for real people like yourself commenting, it's to baffle the bots